Temperature sensors

[Dave Plowman (News)]

Been playing with air temperature sensors on my old Rover - for the fuel injection which is MegaSquirt aftermarket. Which can be programmed for pretty well any sensor.

I want to re-site it, but need to use the old one as a blank to the trunking, and couldn't find an identical one. Although both are meant to be the same electrically - but aren't.

However, here's the point.

If I connect them to my Fluke DVM set to temperature, both read exactly the same - room temperature.

If I measure the resistance, the original is 3K4, the new one 2K0, both at room temperature which the Fluke says is 22.7C on both of them and its own sensor.

MegaSquirt reports ambient correctly with the old sensor but reads approx

8C high with the new one. Not a problem as I can enter the values of the new one in the MS software at a few spot temperatures and it will interpolate the rest.

But I'm curious to know just how the Fluke copes with both without adjustment?

[Davey]

I know nothing about car temp. sensors, but will try to help if I can. Are they thermocouples? RTDs? Thermistors?

[Davey]

You might find better help in uk.rec.cars.maintenance.

[Dave Plowman (News)]

Negative temperature coefficient thermistors, IIRC.

[Dave Plowman (News)]

I doubt it. ;-)

[Andrew Gabriel]

Is it initially adjusting internally to give same temperature as it's internal sensor?

[Davey]

Thermistors are the ones I don't know anything about!

Sorry.

[Dave Plowman (News)]

Turns out it was a big red herring. The Fluke reads about 22C over a wide range of resistances when switched to temperature. But isn't actually giving the temperature as I discovered when I put the sensor in the freezer compartment to get the low temperature resistance. The Fluke still read 22C from it - but -5C using its own sensor. I think the Fluke probably uses a thermocouple. That reads only a couple of ohms resistance.

Anyway, the calibration is done and it now reads exactly the same as the coolant sensor on a cold engine.